BUG: Soft Lockup - CPU#0 stuck for 22s! [qemu-system-x86: 31867]

[Prem Karat]

Environment:
-------------------
        * Upstream git version: qemu-kvm-1.1-rc2-4-g3fd9fed
        * Host Kernel: Mainline Kernel - 3.4.0 x86_64 GNU/Linux (Arch: x86_64)
        * CPU model: Intel(R) Xeon(R) CPU X5570 @ 2.93GHz
        * Guest OS: Red Hat Enterprise Linux Server release 6.2
        * Guest Kernel: 2.6.32-220.el6.x86_64
        * Qemu-command line:
/usr/local/bin/qemu-system-x86_64 -name 'vm1' -nodefaults -monitor unix:'/tmp/monitor-humanmonitor1-20120525-214210-Zua6',server,nowait -serial unix:'/tmp/serial-20120525-214210-Zua6',server,nowait -device ich9-usb-uhci1,id=usb1 -drive file='/tmp/kvm_autotest_root/images/rhel62-64.qcow2',index=0,if=ide,cache=none -device rtl8139,netdev=idvVySvg,mac='9a:6d:16:b9:b5:06',id='idiX1NmG' -netdev tap,id=idvVySvg,fd=21 -m 7198 -smp 2 -device usb-tablet,id=usb-tablet1,bus=usb1.0 -vnc :0 -vga std

The qemu is started through autotest.

Description:
-----------------

While running the cgroup test through autotest, the host was hung and was not responding. When viewed through serial console, found the error "BUG: Soft lockup" error as attached in the screenshot 1.

There are no errors displayed in /var/log/messages (no call trace) and in dmesg.*
There is a call trace seen in serial console, which is show in screenshot 2.

Steps to reproduce:
----------------------------
Currently am not able to consistently reproduce this error. However when I tried to reproduce it again by running the cgroup test, found another error from syslogd as shown below

"Message from syslogd@phx3 at May 25 21:56:04 ...
  kernel:Kernel panic - not syncing: Watchdog detected hard LOCKUP on cpu 3"

So this time I got a hard Lockup error. Attached is the screenshot of the same. (screenshot-3, see the message at the bottom of the screen). This time the cgroup test had completed.

Please let me know if you require more info on this.

-prem

[Thomas Huth (th-huth)]

Thanks for the bug report, but please report kernel bugs in the kernel bug tracker, not in the QEMU bug tracker (see http://www.linux-kvm.org/page/Bugs for details). So if the problem still persists with recent kernels, you should open a ticket there instead.